A data processing method, device and equipment applied to audit system

By obtaining and formatting unstructured data of different types of financial systems and comparing and presenting it, the problems of low audit efficiency and high cost are solved, and rapid review and efficient audit are achieved.

CN118132673BActive Publication Date: 2024-11-26SHANDONG DEZHOU TOBACCO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410225013.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-29
Publication Date
2024-11-26
Estimated Expiration
2044-02-29

AI Technical Summary

Technical Problem

In audit scenarios, the existing technology has problems of low audit efficiency and high audit cost, mainly due to incomplete data collection and long manual auditing.

Method used

By obtaining unstructured raw data of different types of financial systems, converting and comparing formats, comparing preset functions with payment standards or audit conditions, and visually displaying them on the display interface to improve audit efficiency and accuracy .

Benefits of technology

It realizes rapid review and verification of the audit process, reduces time and labor costs, and improves the accuracy and efficiency of audit work.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118132673B_ABST
    Figure CN118132673B_ABST
Patent Text Reader

Abstract

The present invention provides a data processing method, device and equipment applied to an audit system, which relates to the field of computer information processing technology. The method includes: obtaining multiple unstructured original data of at least two different types of financial systems; performing format conversion on the first type of unstructured original data of the first type of financial system of the at least two different types to obtain first conversion result data; performing format conversion on the second type of unstructured original data of the second type of financial system of the at least two different types to obtain second conversion result data; comparing the first conversion result data and the second conversion result data with the payment standard or the audit condition respectively through a preset function to obtain a comparison result; and visually displaying the comparison result on a display interface. The solution of the present invention realizes rapid review and verification of the audit process, reduces time and labor costs, and improves the accuracy and efficiency of the audit work.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of computer information processing, and in particular to a data processing method, device and equipment applied to an audit system. Background Art

[0002] In the audit scenario, the audit work is mainly based on sampling and static auditing, which requires on-site evidence collection and review of paper evidence one by one. In addition, due to the low data concentration, it is very easy for data collection to be incomplete, and manual auditing takes a long time. Therefore, it takes a lot of time to search, collect, calculate, and compare, which has the disadvantages of low audit efficiency and high audit cost. Summary of the invention

[0003] The present invention provides a data processing method, device and equipment applied to an audit system, which solves the technical problems of low audit efficiency and high audit cost in the existing audit process.

[0004] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0005] An embodiment of the present invention provides a data processing method applied to an audit system, comprising:

[0006] Obtain multiple unstructured raw data from at least two different types of financial systems;

[0007] Performing format conversion on a first type of unstructured original data of a first type of financial system among the at least two different types to obtain first conversion result data;

[0008] Performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data;

[0009] By using a preset function, the first conversion result data and the second conversion result data are respectively compared with the payment standard or the audit condition to obtain a comparison result;

[0010] The comparison results are visually displayed on a display interface.

[0011] Optionally, obtain multiple unstructured raw data from at least two different types of financial systems, including:

[0012] Determine the specific type of data required based on the audit objectives and requirements;

[0013] Obtain the necessary permissions and credentials to access financial systems, depending on the specific type of data required;

[0014] Once permissions and credentials are obtained, establish connections with at least two financial systems;

[0015] Once connected, obtain the specific location of the unstructured data source based on the data structure of each financial system;

[0016] Get multiple unstructured raw data from different types of financial systems based on the specific location of the unstructured data source.

[0017] Optionally, performing format conversion on the first type of unstructured original data of the first type of financial system among the at least two different types to obtain first conversion result data includes:

[0018] Obtaining a first text structure feature quantity of the first type of unstructured original data;

[0019] Performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector;

[0020] Constructing a query control function according to the first digitized feature vector by a first preset method;

[0021] According to the query control function, the first type of unstructured raw data file number is processed by a second preset method to extract key information from the first type of unstructured raw data;

[0022] The key information in the first type of unstructured original data is converted into a preset format to obtain first conversion result data.

[0023] Optionally, performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data includes:

[0024] Obtaining a second text structure feature quantity of the second type of unstructured original data;

[0025] Performing windowing processing on the second text structure feature quantity to obtain a second numerical feature vector;

[0026] According to the second digitized feature vector, constructing a query control function by a first preset method;

[0027] According to the query control function, the second type of unstructured raw data file number is processed by a second preset method to extract key information from the second type of unstructured raw data;

[0028] The key information in the second type of unstructured original data is converted into a preset format to obtain second conversion result data.

[0029] Optionally, obtaining a first text structure feature of the first type of unstructured original data includes:

[0030] Read the first type of unstructured raw data stored in the financial system;

[0031] Analyze the first type of unstructured raw data to obtain the overall situation and structural characteristics of the first type of unstructured raw data;

[0032] Analyze the first text structure according to the overall situation and structural characteristics of the first type of unstructured original data to obtain the first text structure analysis result;

[0033] Defining a first text structure feature according to the first text structure analysis result;

[0034] A defined first text structure feature quantity is extracted from the first type of unstructured raw data.

[0035] Optionally, performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector includes:

[0036] Performing characteristic analysis on the first text structure feature quantity to obtain key attributes of the first text structure feature quantity;

[0037] Determining the window size according to the key attributes of the first text structure feature quantity and the detail level of the feature to be captured;

[0038] Dividing the first text structure feature into a plurality of windows according to the window size;

[0039] The internal text structure feature quantity of each divided window is extracted to obtain the first numerical feature vector.

[0040] Optionally, constructing a query control function according to the first digitized feature vector by a first preset method includes:

[0041] Obtaining the dimension and complexity of the first digitized feature vector;

[0042] Determine the number of potential solutions that will be generated based on the dimensionality and complexity of the feature vector;

[0043] Adjustments are made to the pheromone levels and randomness of the solutions to generate new solutions;

[0044] Evaluate the performance of each solution and obtain the evaluation results;

[0045] The pheromone level is updated according to the evaluation results until the preset conditions are met, and the final solution currently determined is output as the query control function.

[0046] The present invention also provides a data processing device applied to an audit system, comprising:

[0047] An acquisition module, used for acquiring a plurality of unstructured raw data of at least two different types of financial systems;

[0048] A processing module is used to perform format conversion on a first type of unstructured original data of a first type of financial system among the at least two different types to obtain first conversion result data; perform format conversion on a second type of unstructured original data of a second type of financial system among the at least two different types to obtain second conversion result data; compare the first conversion result data and the second conversion result data with the payment standard or the audit condition respectively through a preset function to obtain a comparison result; and visualize the comparison result on a display interface.

[0049] The present invention also provides a computing device, comprising: a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction runs on a computer, the computer executes the above-mentioned data processing method applied to an audit system.

[0050] The present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enables the computer to execute the above-mentioned data processing method applied to an audit system.

[0051] The above solution of the present invention includes at least the following beneficial effects:

[0052] The data processing method applied to the audit system of the present invention obtains multiple unstructured original data of at least two different types of financial systems; performs format conversion on the first type of unstructured original data of the first type of financial system among the at least two different types to obtain first conversion result data; performs format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data; compares the first conversion result data and the second conversion result data with the payment standard or audit condition respectively through a preset function to obtain a comparison result; and visualizes the comparison result on a display interface. Rapid review and verification of the audit process is achieved, time and labor costs are reduced, and the accuracy and efficiency of the audit work are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of a data processing method applied to an audit system of the present invention;

[0054] Figure 2 It is a schematic diagram of the structure of the data processing device applied to the audit system of the present invention. DETAILED DESCRIPTION

[0055] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to enable the scope of the present invention to be fully communicated to those skilled in the art.

[0056] like Figure 1 As shown, an embodiment of the present invention provides a data processing method applied to an audit system, comprising:

[0057] Step 11, obtaining a plurality of unstructured raw data of at least two different types of financial systems;

[0058] Step 12, performing format conversion on the first type of unstructured original data of the first type of financial system among the at least two different types to obtain first conversion result data;

[0059] Step 13, performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data;

[0060] Step 14, using a preset function, respectively comparing the first conversion result data and the second conversion result data with the payment standard or the audit condition to obtain a comparison result;

[0061] Step 15: Visually display the comparison results on a display interface.

[0062] In this embodiment, the data processing method applied to the audit system can integrate unstructured raw data from multiple different types of financial systems. By obtaining unstructured raw data from at least two different types of financial systems and performing format conversion, these data can be integrated together to facilitate subsequent data analysis and comparison. By comparing the payment standards or audit conditions, by using the vlookup function (vertical search function), the first conversion result data and the second conversion result data are compared with the payment standards or audit conditions respectively, which can quickly discover possible problems or abnormal situations, such as the payment amount not meeting the standard or the existence of duplicate payments, etc. At the same time, the comparison results can be intuitively displayed in the form of charts, tables, etc., to help auditors better understand and analyze the data and discover potential risks and problems. The audit system can process and analyze large amounts of financial data more efficiently and improve the accuracy and efficiency of the audit work. At the same time, the visual display can help auditors quickly understand the data and provide decision-making support and communication tools.

[0063] In an optional embodiment of the present invention, step 11 may include:

[0064] Step 111, determine the specific type of data required based on the audit purpose and requirements;

[0065] Step 112, obtaining the permissions and credentials required to access the financial system according to the specific type of data required;

[0066] Step 113, after obtaining the authority and credentials, establishing connections with at least two financial systems;

[0067] Step 114, after the connection is made, the specific location of the unstructured data source is obtained according to the data structure of each financial system;

[0068] Step 115 , obtaining a plurality of unstructured raw data of different types of financial systems according to the specific location of the unstructured data source.

[0069] In this embodiment, step 111 is the starting point of the audit process. The auditor first needs to clarify the audit objectives, such as whether to check the accuracy and compliance of financial statements or to identify potential fraud. Based on these objectives, it can be determined what specific types of data need to be collected. These data may include transaction records, account balances, customer or supplier information, etc. Therefore, the present invention can ensure that the collected data is closely related to the audit objectives, improve audit efficiency, avoid collecting unnecessary data, and save time and resources. After determining the required data type, it is necessary to obtain appropriate system access rights, which involves coordinating with the IT department or system administrator to ensure that the auditor has sufficient rights to access, view and export the required data. These rights and credentials may include user names, passwords, security tokens or specific system roles, which can ensure that only authorized personnel can access sensitive data. Once the necessary access rights are obtained, the auditor Once the auditor has connected to the financial system, he or she can start to establish a connection with the financial system, which usually involves the use of specialized software tools or platforms that are compatible with and can exchange data with various financial systems; establishing connections with multiple systems can ensure that auditors can obtain comprehensive financial data, thereby conducting more accurate audit analysis. By collecting data from multiple systems, a more comprehensive financial view can be obtained, which can adapt to different financial system environments without relying on a single data source; after connecting to the financial system, it is necessary to have an in-depth understanding of the data structure of each system, including identifying which data is stored in an unstructured form (such as text files, e-mails, images, etc.), and determining the specific location of this data in the system. Understanding the data structure is crucial for subsequent data extraction, conversion, and analysis. The present invention can accurately locate unstructured data and ensure the accuracy of audit analysis. By pre-analyzing the data structure, the data extraction process can be optimized and errors and delays in subsequent processing can be reduced.

[0070] In an optional embodiment of the present invention, step 12 may include:

[0071] Step 121, obtaining a first text structure feature value of the first type of unstructured original data;

[0072] Step 122, performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector;

[0073] Step 123, constructing a query control function by a first preset method according to the first digitized feature vector;

[0074] Step 124, according to the query control function, the first type of unstructured raw data file number is processed by a second preset method to extract key information from the first type of unstructured raw data;

[0075] Step 125: convert key information in the first type of unstructured original data into a preset format to obtain first conversion result data.

[0076] In this embodiment, when obtaining the first text structure feature quantity of the first category of unstructured original data, it is first necessary to obtain the first text structure feature quantity of the first category of unstructured original data. The first text structure feature quantity may include information such as keywords, paragraph structure, and title, which is used for subsequent data processing and analysis. The obtained first text structure feature quantity is subjected to windowing processing. Windowing processing refers to segmenting the first text structure feature quantity according to a certain window size so that each window can be further processed, which can better focus on the local features of the text and improve the accuracy of data processing. According to the digitized feature vector, a query control function is constructed by an ant colony optimization method. The ant colony optimization method is an optimization algorithm that simulates the foraging behavior of an ant colony. The query control function can be constructed according to the digitized feature vector to locate and extract the first category of unstructured data. The key information in the structured raw data can be processed quickly and accurately in the case of large amounts of data. According to the query control function, the slicing method in the Python (computer programming language) development language is used to process the file numbers of the first type of unstructured raw data. By using the slicing method in the Python development language, the file numbers in the first type of unstructured raw data can be processed according to the query control function to extract key information. Finally, the extracted key information is converted into a preset format such as Excel format, so that auditors can more conveniently perform data analysis and visualization. Format conversion of the first type of unstructured raw data and extraction of key information therein will help auditors better understand and analyze data, thereby improving the accuracy and efficiency of audit work.

[0077] In an optional embodiment of the present invention, step 13 may include:

[0078] Step 131, obtaining a second text structure feature value of the second type of unstructured original data;

[0079] Step 132, performing windowing processing on the second text structure feature quantity to obtain a second numerical feature vector;

[0080] Step 133, constructing a query control function by a first preset method according to the second digitized feature vector;

[0081] Step 134, according to the query control function, the second type of unstructured raw data file number is processed by a second preset method to extract key information from the second type of unstructured raw data;

[0082] Step 135: convert the key information in the second type of unstructured original data into a preset format to obtain second conversion result data.

[0083] In this embodiment, the specific process of format conversion of the second type of unstructured original data of the second type of financial system among the at least two different types to obtain the second conversion result data is the same as step 12, both of which use the ant colony optimization method to construct a query control function, and according to the query control function, use the slicing method in the Python (computer programming language) development language to process the file number of the second type of unstructured original data, so as to obtain the key information in the second type of unstructured original data, and convert the key information in the second type of unstructured original data into a preset format to obtain the second conversion result data, which helps auditors to better understand and analyze the data, thereby improving the accuracy and efficiency of the audit work and meeting the requirements of the audit system for various financial system data.

[0084] In an optional embodiment of the present invention, step 121 may include:

[0085] Step 1211, reading the first type of unstructured raw data stored in the financial system;

[0086] Step 1212, analyzing the first type of unstructured original data to obtain the overall situation and structural characteristics of the first type of unstructured original data;

[0087] Step 1213, analyzing the first text structure according to the overall situation and structural characteristics of the first type of unstructured original data, and obtaining a first text structure analysis result;

[0088] Step 1214, defining a first text structure feature according to the first text structure analysis result;

[0089] Step 1215: extract the defined first text structure feature from the first type of unstructured original data.

[0090] In this embodiment, firstly, the first type of unstructured raw data can be read from the connected financial system through a data processing tool or platform. These data may include documents in various formats, such as PDF format documents, Word format documents, emails, pictures, videos, etc. They contain important financial and business information, but are usually not stored in the form of traditional tables or databases. The present invention can effectively obtain unstructured data from complex financial systems, provide a basis for subsequent analysis, ensure that the analysis covers all relevant unstructured data, and improve the integrity of the audit; by performing a preliminary analysis on the obtained unstructured data to understand the overall situation of the data, such as data volume, data type, data format, etc., and identify the structural characteristics of the data, such as titles, paragraphs, tables, lists, etc. in the document; the present invention can help auditors better understand the content and structure of unstructured data, prepare for subsequent more in-depth analysis, and optimize data processing and analysis strategies based on the structural characteristics of the data; the data processing method applied to the audit system can utilize natural language processing (NLP) technology or other related The present invention uses a related tool to conduct an in-depth analysis of the text content in the unstructured data, including identifying the key elements in the text, understanding the logical structure of the text, extracting entities and relationships in the text, etc., which can extract the key information in the text, provide in-depth business insights for auditors, and process a large amount of text data through automated tools to improve audit efficiency; based on the results of the text structure analysis, a series of text structure feature quantities are defined, which can be the length of the text, the frequency of keywords, the number of occurrences of specific sentence patterns, the emotional tendency of the text, etc., which are used to quantify the structure and content characteristics of the text; the present invention converts text data into quantifiable feature quantities, which is convenient for statistical analysis and pattern recognition, and defines a unified feature quantity standard, so that the comparison and analysis between different texts are more objective and accurate; step 1215 can extract the previously defined text structure feature quantities from the original text data by using data processing tools or platforms, and these feature quantities will be used for subsequent data analysis, model training and audit decision-making. By extracting key feature quantities, the present invention can reduce the impact of data noise and irrelevant information on audit results and improve the accuracy of audit.

[0091] In an optional embodiment of the present invention, step 122 may include:

[0092] Step 1221, performing characteristic analysis on the first text structure feature quantity to obtain key attributes of the first text structure feature quantity;

[0093] Step 1222, determining the window size according to the key attributes of the first text structure feature quantity and the detail level of the feature to be captured;

[0094] Step 1223, dividing the first text structure feature into a plurality of windows according to the window size;

[0095] Step 1224: extract the internal text structure feature quantity of each divided window to obtain a first numerical feature vector.

[0096] In this embodiment, a characteristic analysis is performed on the first text structure feature, including identifying the statistical characteristics of the feature (such as mean, variance, distribution, etc.), time series characteristics (such as periodicity, trend, etc.) and the correlation between them. Through such analysis, it is possible to determine which attributes are critical to subsequent analysis and modeling, and a deeper understanding of the nature and behavior of the text structure feature helps to select more appropriate processing and analysis methods. By identifying key attributes, irrelevant or redundant features can be removed, the dimension and noise of the data can be reduced, and the accuracy and efficiency of subsequent analysis can be improved; the selection of window size is very important for windowing processing. If the window is too small, it may not be able to capture enough feature details, while if the window is too large, it may cause information loss or low computational efficiency. Therefore, it is necessary to determine a suitable window size based on the key attributes of the feature and the analysis target; by reasonably selecting the window size, a balance can be achieved between computational efficiency and feature detail capture. Different texts and analysis targets may require different window sizes. This method provides flexibility and adaptability; once the window size is determined, the text structure features can be divided into multiple windows according to time series or other orders, and each window contains feature data for a period of time or a part of the text; windowing allows the characteristics of text data to be analyzed locally, which helps capture local patterns and changes. By dividing the data into smaller windows, the complexity of the analysis can be reduced, making the processing of large-scale text data more feasible; within each window, the previously defined text structure features are extracted and combined into a numerical feature vector, which can be used as the input of subsequent machine learning models for training and prediction; converting text data into numerical feature vectors is a common practice in machine learning, which enables text data from different sources and formats to be processed by the model in a unified manner. By converting text data into feature vectors, the feature patterns of text data can be more easily understood and analyzed, which helps to improve the interpretability of the model.

[0097] In an optional embodiment of the present invention, step 123 may include:

[0098] Step 1231, obtaining the dimension and complexity of the first digitized feature vector;

[0099] Step 1232, determining the number of potential solutions to be generated based on the dimension and complexity of the feature vector;

[0100] Step 1233, adjusting according to the pheromone level and randomness of the solution to generate a new solution;

[0101] Step 1234, evaluating the performance of each solution to obtain an evaluation result;

[0102] Step 1235, updating the pheromone level according to the evaluation result;

[0103] Step 1236, repeat the operation until the preset conditions are met, and output the final solution currently determined as the query control function. Specifically, each iteration generates a new solution based on the current pheromone level and solution set, and after each iteration, the final solution is determined as the current optimal solution; when the termination condition is met, stop the iteration, and output the final solution currently determined as the final instance of the query control function.

[0104] In this embodiment, the dimension of the first numerical feature vector, that is, the number of features or the length of features, is first analyzed. At the same time, the complexity of these features, such as the correlation between features, nonlinear relationships, etc., needs to be evaluated. These analyses help to understand the structure and characteristics of the data and provide a basis for the subsequent optimization process. By obtaining the dimension and complexity of the feature vector, the structure and characteristics of the data can be more deeply understood. This information can help set appropriate algorithm parameters, such as the number of solutions, the initial level of pheromones, etc. The number of potential solutions is an important parameter in the ant colony optimization algorithm. According to the dimension and complexity of the feature vector, the size of the solution space to be explored can be estimated, so as to determine a suitable number of potential solutions. This number should be large enough to cover the possible solution space, but should not be too large to avoid unnecessary computational overhead. By reasonably setting the number of potential solutions, a balance can be achieved between exploring new solutions and using known information, avoiding the generation of too many solutions, saving computing resources, and improving the efficiency of the algorithm. In the ant colony optimization algorithm, new solutions are generated by simulating the movement of ants in the solution space. Ants will choose the direction of movement or select features based on the pheromone level and randomness; the direction or feature with a higher pheromone level is more likely to be selected, but randomness also ensures that the algorithm can explore new and possibly better solutions; by introducing randomness, the algorithm can jump out of the local optimal solution and find the global optimal solution. The pheromone level will be dynamically adjusted according to the performance of the solution, so that the algorithm can adaptively optimize the search process; define an evaluation function to measure the performance of each solution. This evaluation function should be able to reflect the degree of optimization of the solution for the query control function. By evaluating the performance of each solution, an evaluation result can be obtained to compare the advantages and disadvantages of different solutions. Through the evaluation results, the performance of different solutions can be quantitatively compared to provide a basis for subsequent selection and update. The evaluation results can also be used to guide the search direction of the algorithm, so that the algorithm can find a better solution faster; in the ant colony optimization algorithm, the pheromone level is updated according to the performance of the solution. Solutions with better performance will release more pheromones to attract more ants to choose these directions or features. In this way, the algorithm will gradually focus on the solution space area with better performance.By increasing the pheromone level of excellent solutions, the algorithm can converge to the global optimal solution more quickly. The pheromone update mechanism enables the algorithm to adaptively adjust the search strategy according to the feedback in the search process. The algorithm will repeat the above steps. Each iteration will generate a new solution based on the current pheromone level and solution set. After each iteration, the algorithm will evaluate the performance of all solutions and select the solution with the best performance as the current optimal solution. This process will continue until the termination condition is met. Through multiple iterations, the algorithm can continuously optimize the solution and improve the performance of the query control function. As the iteration proceeds, the algorithm will gradually approach the global optimal solution and improve the quality of the solution. When the algorithm meets a certain termination condition (such as reaching the maximum number of iterations, the performance of the solution is no longer significantly improved, etc.), the algorithm will stop iterating and output the final solution currently determined. This solution is the final instance of the query control function and can be used for actual query control tasks. By setting appropriate termination conditions, the balance between the running time of the algorithm and the quality of the solution can be controlled. The output final solution can be directly applied to actual query control tasks to improve the efficiency and accuracy of queries.

[0105] The data processing method applied to the audit system described in the present invention is that the data volume in the digital audit system is large, which reduces the data query efficiency and precision rate. Therefore, a digital audit system data fast query method based on ant colony optimization is proposed; a distributed link node following recognition method is used to collect digital audit system data, and the collected data is reorganized by nonlinear samples, and the digital audit system data is reduced in dimension by combining template matching and linear programming design methods to extract data features; on this basis, the abnormal features of the digital audit system data query are judged according to the differences of ant colony individuals, the reliability text structure feature quantity of the data is obtained, and the text structure feature quantity is windowed. According to the processing result, the query control function is constructed by using the ant colony optimization method to obtain the global optimal solution, and the digital audit system data is quickly queried and reviewed, thereby solving the technical problems of low audit efficiency and high audit cost in the existing audit process, realizing automatic comparison of data collected from various existing business systems, early warning of abnormal data, and manual verification of abnormal data by auditors, thereby reducing time and labor costs and improving the accuracy and efficiency of audit work.

[0106] like Figure 2 As shown, an embodiment of the present invention further provides a data processing device 20 applied to an audit system, comprising:

[0107] An acquisition module 21, used for acquiring a plurality of unstructured raw data of at least two different types of financial systems;

[0108] The processing module 22 is used to perform format conversion on a first type of unstructured original data of a first type of financial system among the at least two different types to obtain first conversion result data; perform format conversion on a second type of unstructured original data of a second type of financial system among the at least two different types to obtain second conversion result data; compare the first conversion result data and the second conversion result data with the payment standard or the audit condition respectively through a preset function to obtain a comparison result; and visualize the comparison result on a display interface.

[0109] Optionally, obtain multiple unstructured raw data from at least two different types of financial systems, including:

[0110] Determine the specific type of data required based on the audit objectives and requirements;

[0111] Obtain the necessary permissions and credentials to access financial systems, depending on the specific type of data required;

[0112] Once permissions and credentials are obtained, establish connections with at least two financial systems;

[0113] Once connected, obtain the specific location of the unstructured data source based on the data structure of each financial system;

[0114] Get multiple unstructured raw data from different types of financial systems based on the specific location of the unstructured data source.

[0115] Optionally, performing format conversion on the first type of unstructured original data of the first type of financial system among the at least two different types to obtain first conversion result data includes:

[0116] Obtaining a first text structure feature quantity of the first type of unstructured original data;

[0117] Performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector;

[0118] Constructing a query control function according to the first digitized feature vector by a first preset method;

[0119] According to the query control function, the first type of unstructured raw data file number is processed by a second preset method to extract key information from the first type of unstructured raw data;

[0120] The key information in the first type of unstructured original data is converted into a preset format to obtain first conversion result data.

[0121] Optionally, performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data includes:

[0122] Obtaining a second text structure feature quantity of the second type of unstructured original data;

[0123] Performing windowing processing on the second text structure feature quantity to obtain a second numerical feature vector;

[0124] According to the second digitized feature vector, constructing a query control function by a first preset method;

[0125] According to the query control function, the second type of unstructured raw data file number is processed by a second preset method to extract key information from the second type of unstructured raw data;

[0126] The key information in the second type of unstructured original data is converted into a preset format to obtain second conversion result data.

[0127] Optionally, obtaining a first text structure feature of the first type of unstructured original data includes:

[0128] Read the first type of unstructured raw data stored in the financial system;

[0129] Analyze the first type of unstructured raw data to obtain the overall situation and structural characteristics of the first type of unstructured raw data;

[0130] Analyze the first text structure according to the overall situation and structural characteristics of the first type of unstructured original data to obtain the first text structure analysis result;

[0131] Defining a first text structure feature according to the first text structure analysis result;

[0132] A defined first text structure feature quantity is extracted from the first type of unstructured raw data.

[0133] Optionally, performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector includes:

[0134] Performing characteristic analysis on the first text structure feature quantity to obtain key attributes of the first text structure feature quantity;

[0135] Determining the window size according to the key attributes of the first text structure feature quantity and the detail level of the feature to be captured;

[0136] Dividing the first text structure feature into a plurality of windows according to the window size;

[0137] The internal text structure feature quantity of each divided window is extracted to obtain the first numerical feature vector.

[0138] Optionally, constructing a query control function according to the first digitized feature vector by a first preset method includes:

[0139] Obtaining the dimension and complexity of the first digitized feature vector;

[0140] Determine the number of potential solutions that will be generated based on the dimensionality and complexity of the feature vector;

[0141] Adjustments are made to the pheromone levels and randomness of the solutions to generate new solutions;

[0142] Evaluate the performance of each solution and obtain the evaluation results;

[0143] The pheromone level is updated according to the evaluation results until the preset conditions are met, and the final solution currently determined is output as the query control function.

[0144] It should be noted that the device is a device corresponding to the above-mentioned data processing method applied to the audit system. All implementation methods in the above-mentioned method are applicable to the embodiments of the device and can achieve the same technical effects.

[0145] An embodiment of the present invention also provides a computing device, comprising: a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the data processing method as described for application to an audit system.

[0146] The embodiment of the present invention further provides a computer-readable storage medium, including instructions, which, when executed on a computer, enable the computer to execute the data processing method applied to an audit system as described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0147] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0148] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0149] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0150] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0151] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0152] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0153] In addition, it should be noted that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understood that all or any steps or components of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in hardware, firmware, software or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0154] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0155] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A data processing method applied to an audit system, characterized in that: include: Obtain multiple unstructured raw data from at least two different types of financial systems; Performing format conversion on a first type of unstructured original data of a first type of financial system among the at least two different types to obtain first conversion result data; Performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data; By using a preset function, the first conversion result data and the second conversion result data are respectively compared with the payment standard or the audit condition to obtain a comparison result; Visually displaying the comparison results on a display interface; The first type of unstructured original data of the first type of financial system of the at least two different types is format converted to obtain first conversion result data, including: Obtaining a first text structure feature quantity of the first type of unstructured original data; Performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector; Constructing a query control function according to the first digitized feature vector by a first preset method; According to the query control function, the first type of unstructured raw data file number is processed by a second preset method to extract key information from the first type of unstructured raw data; Convert key information in the first type of unstructured raw data into a preset format to obtain first conversion result data; Performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data includes: Obtaining a second text structure feature quantity of the second type of unstructured original data; Performing windowing processing on the second text structure feature quantity to obtain a second numerical feature vector; According to the second digitized feature vector, constructing a query control function by a first preset method; According to the query control function, the second type of unstructured raw data file number is processed by a second preset method to extract key information from the second type of unstructured raw data; Convert key information in the second type of unstructured raw data into a preset format to obtain second conversion result data; The first text structure feature quantity of the first type of unstructured raw data is obtained, including: Read the first type of unstructured raw data stored in the financial system; Analyze the first type of unstructured raw data to obtain the overall situation and structural characteristics of the first type of unstructured raw data; Analyze the first text structure according to the overall situation and structural characteristics of the first type of unstructured original data to obtain the first text structure analysis result; Defining a first text structure feature according to the first text structure analysis result; Extracting a defined first text structure feature quantity from the first type of unstructured raw data; Performing window processing on the first text structure feature quantity to obtain a first numerical feature vector includes: Performing characteristic analysis on the first text structure feature quantity to obtain key attributes of the first text structure feature quantity; Determining the window size according to the key attributes of the first text structure feature quantity and the detail level of the feature to be captured; Dividing the first text structure feature into a plurality of windows according to the window size; Extracting the internal text structure feature quantity of each divided window to obtain a first numerical feature vector; According to the first digitized feature vector, a query control function is constructed by a first preset method, including: Obtaining the dimension and complexity of the first digitized feature vector; Determine the number of potential solutions that will be generated based on the dimensionality and complexity of the feature vector; Adjustments are made to the pheromone levels and randomness of the solutions to generate new solutions; Evaluate the performance of each solution and obtain the evaluation results; Update the pheromone level according to the evaluation results until the preset conditions are met, and output the final solution currently determined as the query control function; The second preset method is a slicing method in a computer programming development language; The first text structure feature includes: keywords, paragraph structure and title information of the first type of unstructured original data; The second text structure feature includes: keywords, paragraph structure and title information of the second type of unstructured original data; The overall situation of the first category of unstructured raw data includes: the data volume, data type and data format of the first category of unstructured raw data; The structural characteristics of the first type of unstructured raw data include: titles, paragraphs, tables, and lists in the documents of the first type of unstructured raw data; The characteristic analysis includes: identifying the statistical characteristics and time series characteristics of the characteristic quantity and the correlation between the statistical characteristics and the time series characteristics; The statistical characteristics include mean, variance and distribution; the time series characteristics include periodicity and trend.

2. The data processing method applied to the audit system according to claim 1 is characterized in that: Obtain multiple unstructured raw data from at least two different types of financial systems, including: Determine the specific type of data required based on the audit objectives and requirements; Obtain the necessary permissions and credentials to access financial systems, depending on the specific type of data required; Once permissions and credentials are obtained, establish connections with at least two financial systems; Once connected, obtain the specific location of the unstructured data source based on the data structure of each financial system; Get multiple unstructured raw data from different types of financial systems based on the specific location of the unstructured data source.

3. A data processing device applied to an audit system, characterized in that: include: An acquisition module, used for acquiring a plurality of unstructured raw data of at least two different types of financial systems; a processing module, configured to perform format conversion on a first type of unstructured original data of a first type of financial system among the at least two different types to obtain first conversion result data; perform format conversion on a second type of unstructured original data of a second type of financial system among the at least two different types to obtain second conversion result data; compare the first conversion result data and the second conversion result data with the payment standard or the audit condition respectively through a preset function to obtain a comparison result; and visually display the comparison result on a display interface; The first type of unstructured original data of the first type of financial system of the at least two different types is format converted to obtain first conversion result data, including: Obtaining a first text structure feature quantity of the first type of unstructured original data; Performing windowing processing on the first text structure feature quantity to obtain a first numerical feature vector; Constructing a query control function according to the first digitized feature vector by a first preset method; According to the query control function, the first type of unstructured raw data file number is processed by a second preset method to extract key information from the first type of unstructured raw data; Convert key information in the first type of unstructured raw data into a preset format to obtain first conversion result data; Performing format conversion on the second type of unstructured original data of the second type of financial system among the at least two different types to obtain second conversion result data includes: Obtaining a second text structure feature quantity of the second type of unstructured original data; Performing windowing processing on the second text structure feature quantity to obtain a second numerical feature vector; According to the second digitized feature vector, constructing a query control function by a first preset method; According to the query control function, the second type of unstructured raw data file number is processed by a second preset method to extract key information from the second type of unstructured raw data; Convert key information in the second type of unstructured raw data into a preset format to obtain second conversion result data; The first text structure feature quantity of the first type of unstructured raw data is obtained, including: Read the first type of unstructured raw data stored in the financial system; Analyze the first type of unstructured raw data to obtain the overall situation and structural characteristics of the first type of unstructured raw data; Analyze the first text structure according to the overall situation and structural characteristics of the first type of unstructured original data to obtain the first text structure analysis result; Defining a first text structure feature according to the first text structure analysis result; Extracting a defined first text structure feature quantity from the first type of unstructured raw data; Performing window processing on the first text structure feature quantity to obtain a first numerical feature vector includes: Performing characteristic analysis on the first text structure feature quantity to obtain key attributes of the first text structure feature quantity; Determining the window size according to the key attributes of the first text structure feature quantity and the detail level of the feature to be captured; Dividing the first text structure feature into a plurality of windows according to the window size; Extracting the internal text structure feature quantity of each divided window to obtain a first numerical feature vector; According to the first digitized feature vector, a query control function is constructed by a first preset method, including: Obtaining the dimension and complexity of the first digitized feature vector; Determine the number of potential solutions that will be generated based on the dimensionality and complexity of the feature vector; Adjustments are made to the pheromone levels and randomness of the solutions to generate new solutions; Evaluate the performance of each solution and obtain the evaluation results; Update the pheromone level according to the evaluation results until the preset conditions are met, and output the final solution currently determined as the query control function; The second preset method is a slicing method in a computer programming development language; The first text structure feature includes: keywords, paragraph structure and title information of the first type of unstructured original data; The second text structure feature includes: keywords, paragraph structure and title information of the second type of unstructured original data; The overall situation of the first category of unstructured raw data includes: the data volume, data type and data format of the first category of unstructured raw data; The structural characteristics of the first type of unstructured raw data include: titles, paragraphs, tables, and lists in the documents of the first type of unstructured raw data; The characteristic analysis includes: identifying the statistical characteristics and time series characteristics of the characteristic quantity and the correlation between the statistical characteristics and the time series characteristics; The statistical characteristics include mean, variance and distribution; the time series characteristics include periodicity and trend.

4. A computing device, characterized in that include: A processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed on a computer, the computer executes the method according to any one of claims 1 to 2.

5. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the method according to any one of claims 1 to 2.